Analog Pre-amplifiers in a Digital Streaming World

[gleeds]

Hi all. I recently made a post on this subject relating to the Westminster, Quest pre-amplifier my firm imports into the US (post# 184 in
Re-imagining Class A Amplification thread).

Today when perusing some of my favorite high-end audio websites, I came across this in depth review by a professional reviewer and frequent contributor on WBF, Mr. Christian Punter. I have the feeling that many members are contemplating this question as more and more members migrate to digital only system. No dissing vinyl. I will forever remain a die-hard fan!

Check it out and weigh in with your own views and experiences on this important subject.

Analog Preamplifiers in a Digital Streaming World

Now that most DACs are equipped with volume control so that they can be connected to a power amp directly, what could there be to gain by switching another component in the signal chain? In other words, when everything on the source end is digital, what is the use of a pure analog preamp? With...

www.hifi-advice.com

 

[antigrunge]

Not sure this proves much of anything: we all agree that what Dacs need is attenuation, not amplification. Digital attenuation is lossy at higher levels of attenuation and therefore anathema.The standard diss on passives rarely applies to digital only systems. Most dacs have output impedances in the low hundreds of Ohms whereas modern power amps nearly always have input impedances exceeding 50 KOhms. So no worries about frquency response unless you listen at -50,db level and even then… Conversely, any pre will add distortion from its amplification circuit which simply isn‘t needed. The best solution in my mind is a DAC with a relay based, i.e. analogue attenuator: no lossy digital attenuation, no added amplification distortion.

 

[godofwealth]

I’ve owned DACs at all price ranges for over 30 years. Haven’t found one yet whose preamp volume control matches what you get from a top quality solid state or tube preamp. And then there’s some of us who have multiple sources (vinyl, tape, FM, CD/SACD player etc.). I don’t see preamps becoming obsolete as long as there’s analog sources to listen to. Even in a purely digital system, eliminating the preamp makes sense if you are trying to minimize space or expense. If you want ultimate transparency, eliminate the DAC entirely by going to a pulse width modulated digital amplifier that does no low level D-to-A conversion.

 

[Gregadd]

I’ve owned DACs at all price ranges for over 30 years. Haven’t found one yet whose preamp volume control matches what you get from a top quality solid state or tube preamp. And then there’s some of us who have multiple sources (vinyl, tape, FM, CD/SACD player etc.). I don’t see preamps becoming obsolete as long as there’s analog sources to listen to. Even in a purely digital system, eliminating the preamp makes sense if you are trying to minimize space or expense. If you want ultimate transparency, eliminate the DAC entirely by going to a pulse width modulated digital amplifier that does no low level D-to-A conversion.

correct.

 

[gleeds]

One DAC with a VC that seems to meet the no preamp challenge is the analog preamp section built into the Lampizator Horizon. Pretty impressive!

 

[godofwealth]

That’s if you can live with a preamp that allows just one analog balanced input (and one unbalanced). My ARC Ref preamp allows 6 XLR inputs — I use three balanced inputs — and 6 unbalanced inputs — I use another 2 of these. If your primary source is digital streaming, the Horizon as a preamp makes sense. But if you’re like me, and use other sources, including vinyl, SACD, other DACs, FM tuner, open reel or cassette tapes etc., then it’s not going to work.

 

[analogsa]

For some reason a separate box with no justifiable tecnical function, placed between dac and power amp, adds a lot of musical goodness if properly designed. Perhaps the separate from the dac PS and the physical remoteness help reduce rf noise.

The majority of decent dacs have volume controls working with more than 32 bits, so bit loss at any volume setting is a bit of an old wives tale. To me, a well implemented digital VC sounds better than anything analogue, let alone a $3 volume control chip used by many.

 

[godofwealth]

I’ve been down this road for over 30 years. I’ve always felt — inserting my left logical brain here — that eliminating the preamp from even a purely digital replay chain should surely result in superior reproduction. It would seem a tautology. But — inserting my right brain here — every time I’ve compared a top class solid state or tube preamp with the direct DAC to amp path, the preamp wins, and by a country mile. I tried this comparison numerous times over the years with different DACs and preamps. Same result every time. I recall in the late 1990s comparing the dCS Elgar Plus — then, the king of the hill in DACs — with feeding its output into an ARC Ref 1. The dCS as a preamp was embarrassingly inferior. To be sure, there was a certain purity to the sound. But the dynamics, the soundstage, and the sheer palpability were all lost. Same result in using a solid state preamp like the Mark Levinson 32. I have besides the Lampi Pacific, 5 other DACs in my house, quite a few have digital volume control, such as the Chord Dave. The Dave on its own sounds much worse. Fortunately my Lampi Pacific has no volume control! One reason I’m loath to “upgrade” to the Horizon. I don’t want a volume control in my DAC if I’m not going to use it. I like the fact that my Pacific has no volume control.

Regarding “32 bits“, ah, you’re getting taken in by the marketing propaganda. Here’s the blunt truth There’s no DAC that can resolve even 24 bits, let alone 32! Check out 25 years of measurements in Stereophile. Not one DAC passes this test. John Atkinson does this comparison on each and every DAC he’s tested over the decades, and he’s tested hundreds. The best DACs maybe resolve 20-21 bits. Of course, even this is completely academic. The world‘s best and most expensive loudspeakers can barely resolve 8 bits in the bass! Check out distortion measurements in ASR. At 96 dB, even high priced JBL professional loudspeakers barely get -50 dB in distortion!

When someone designs a loudspeaker that can resolve 16 bits on a CD, a new era in high fidelity reproduction will be born. That’s really hard. The gap is immense. You’d need to design a loudspeaker that has more than a thousand times less distortion than today’s lowest distortion loudspeakers (e.g., in Stereophile, the Quad 2805 was shown to have -70 dB distortion, and even that’s a far cry from -96 dB of redbook CD, not to mention -140 dB THD of 24-bit digital).

32-bit reproduction? In your dreams. That’s not technologically achievable with today’s materials. Maybe quantum DACs with new type of transistor technology and an entirely novel way to design loudspeakers that no one yet knows how to do will get us there. Not in my lifetime.

 

[antigrunge]

This is obviously ground that has been furrowed many times. There seems no debate that dacs require attenuation, not amplification. The argument against passives of impedance mismatches doesn’t apply to 99% of modern dac and power amp designs. Conversely no argument that I am aware of has been proposed why actives with their demonstrable additional distortion present advantages over passive attenuation. This surely merits a proper designer’s attention, doesn’t it? (Obviously assuming high quality short cable connections)

 

[Chops]

Precisely why I purchased the Quad Artera Pre preamp. It's fully analogue, has a very good MM/MC phono stage, though only a single balanced XLR input and output, which is all I need. However, it also has a single unbalanced output as well as a single "fixed" unbalanced output which is perfect to feed my analogue side of the system (cassette decks and reel-to-reel deck). Unfortunately, it doesn't have an actual "Tape Loop", but I make due with the dbx 400x in the chain which works perfectly for my needs. As for the Artera tone controls and such, I don't touch them. Everything is left flat.

The Artera replaced my very much loved Schiit Freya+ that I used in buffered mode as well as in tube mode with various NOS tubes. I have to say, as much as I loved that preamp and how good it sounded (very involving, wide and deep soundstage, great detail and a fair amount of warmth), it does not posses the dynamics, punch, pure naturalness, air and refinement that the Quad Artera Pre has. The soundstage is even a bit wider and deeper with better layering.

No matter if I'm listening at whisper levels late at night or cranking out some tunes in the middle of the day, the refinement, punch and dynamics remain. In fact, this is the only preamp that I have owned that retains its sound signature at any volume level. Virtually every preamp I have owned in the past would suffer from reduced bass output at low volumes (a.k.a. - late night listening), and at the other extreme, get a bit bright or harsh when pushed to higher volume levels. It simply just gets quieter or louder, that's it. No other change in sound.

I only wish this preamp had more exposure in the media. There's very little in the way or reviews on this preamp, and I've only come across two or three others that own this preamp, and they too love it.

 

[audiobomber]

> when the volume is controlled in the digital domain, bits are discarded, which does reduce the resolution.

As others have mentioned in this thread, that is not an issue with a 32-bit DAC. Specifically, each bit resolves 6dB of dynamic range. Decreasing the level from 32 bits to 24 bits means you can drop the volume by 48dB before losing a single bit. But an audio system can't use anywhere close to 24 bits, which is 144dB of dynamic range. If we use a very high target of 120dB of DR, we can drop the resolution to 20 bits, a volume level of -72dB without dropping actual resolution. My speakers are rated 91dB@2.83V. I typically listen at -24dB.

The analog output stage determines whether a DAC will be a good preamp. Not all DAC's are up to the task. Especially in a system like mine, with a combined output impedance just above 4k?.

 

[Kingrex]

If I had a condo with limited space, a DAC with volume would be nice. Then again, I would purchase an integrated amp with built in streaming.

And that begs a question, of all the integrated amps with a built in DAC, do any bypass the preamp section, or not have one at all, and operate the volume in the digital domain?

I would think a sonic gain that required far less parts would be the path these manufacturers would go. But I believe most all invest in an active preamp section to attenuate the volume.

 

[Macattack]

Hey Rex, My Bricasti M21 DAC/Pre operates that way. If I were to set its volume to max it automatically bypasses its analog preamp section.

This is a topic of interest to me. I have been using my M21 direct to my amps for a few years and don’t feel like I am missing anything. But that bug is in my head so I will be trying a VAC Master tube preamp in the next few weeks. I did audition a Bricasti M20 preamp a year ago and other than higher gain, I didn’t find a compelling difference or improvement in SQ. Others have.

 

[Holmz]

This is obviously ground that has been furrowed many times. There seems no debate that dacs require attenuation, not amplification. The argument against passives of impedance mismatches doesn’t apply to 99% of modern dac and power amp designs. Conversely no argument that I am aware of has been proposed why actives with their demonstrable additional distortion present advantages over passive attenuation. This surely merits a proper designer’s attention, doesn’t it? (Obviously assuming high quality short cable connections)

^OK^, Here goes then:

I suspect that most DACS have a relatively low output impedance…?

But a higher output impedance into a low input impedance of a power amp is where the danger is at.
At high volume the output impedance of the passive-pre is whatever the output impedance of the input device is.
But at low volume, then output impedance increases.

Not a problem when the thing is running on the loud side, but highly sensitive speakers (e.g. horns), and a low listening level can get very high amounts of attenuation.
And then if the amp(s) are a low input impedance, it gets to be a cable challenge.

With an active preamp, the output impedance is always the same, and much lower than an attenuated passive-pre.

 

[audiobomber]

With an active preamp, the output impedance is always the same, and much lower than an attenuated passive-pre.

I agree, and to belabor the point:

With an active preamp or DAC/preamp, the output impedance is always the same, and much lower than an attenuated passive-pre.

 

[BillK]

It’s also all system-dependent.

When I had a Mark Levinson No. 38S (later a No. 380S) my Wadia S7i always sounded better running through it in SSP pass-through mode and using the volume control on the Wadia.

When I upgraded to an Ayre KX-R (and later KX-R Twenty) it sounded better with the Wadia at full volume and letting the Ayre handle volume control.